Really simple crossover question.

[crazybastard]

At the risk of exposing myself as an absolute novice, here goes. If I build a 2 way speaker using 2 8ohm speakers and a parallel circuit, would I effectively have about a 4ohm speaker?

[cjd]

assuming you're talking 8ohm tweeter + 8ohm mid-woofer:

no

you'd have an 8ohm speaker

[Hdale85]

Does your load go by your woofer section? I'm a little curious about this as well heh

[Brian Walter]

No, the load does not go by the woofer, but rather by the combined impedance of the woofer, tweeter and the crossover. But even then, the impedance may be 6 ohms over a good protion of the response and still be considered an 8 ohm impedance.

[crazybastard]

Thanks for the response. So if I wired two 8ohm drivers in parallel, no crossover, the overall resistance is halved right? So in a single first order crossover in parallel, there's just one cap and one coil. There's enough resistance to bring what is a 4ohm speaker up to 8ohm? Is this relationship linear with increased cap value and coil value?

[Dennis H]

Once you get past the big picture idea of how crossovers work, you need to model the impedance and SPL in software. Big picture: the crossover cuts highs or lows by increasing the impedance at those frequencies so the two drivers are never truly in parallel at any given frequency. Pesky details: an "8 ohm" driver may have an impedance that varies between 5 ohms and 30 ohms at different frequencies. How a particular crossover config will react with that is complicated enough that you really need crossover design software along with accurate SPL and impedance curves of the drivers to see what's going on. Textbook formulas for SPL and impedance are pretty worthless in the real world.

[1Michael]

"So if I wired two 8ohm drivers in parallel, no crossover, the overall resistance is halved right"

No, its doubled. In series halves the Ohms.

[cjd]

Thanks for the response. So if I wired two 8ohm drivers in parallel, no crossover, the overall resistance is halved right? So in a single first order crossover in parallel, there's just one cap and one coil. There's enough resistance to bring what is a 4ohm speaker up to 8ohm? Is this relationship linear with increased cap value and coil value?

That's not how it works. You wire the drivers to the crossover, and the two networks are wired in parallel. A capacitor has a rising impedance as frequency drops, with a corner at some specific point. So what happens is you add series impedance that keeps going up as frequency goes down (assuming a series cap). Thus, at 20Hz a cap you would normally put on a tweeter has a very very high impedance. Basically, when you put that in parallel with something, the net impedance is equal to the lowest impedance (i.e. 8ohm). Flip that around for inductors (rising impedance as frequency rises)...

C

[cjd]

"So if I wired two 8ohm drivers in parallel, no crossover, the overall resistance is halved right"

No, its doubled. In series halves the Ohms.

noooo, in parallel = halved (if you've got two equal impedance items wired in parallel). Series doubles.

C

[crazybastard]

Once you get past the big picture idea of how crossovers work, you need to model the impedance and SPL in software. Big picture: the crossover cuts highs or lows by increasing the impedance at those frequencies so the two drivers are never truly in parallel at any given frequency. Pesky details: an "8 ohm" driver may have an impedance that varies between 5 ohms and 30 ohms at different frequencies. How a particular crossover config will react with that is complicated enough that you really need crossover design software along with accurate SPL and impedance curves of the drivers to see what's going on. Textbook formulas for SPL and impedance are pretty worthless in the real world.

Most drivers come with an impedance graph. Can you go by this graph or do you measure your drivers individually. I have a pair of Hippi@#@ something tweeters, and I heard that they were tested and matched individually before leaving the factory. However, most other manufacturers publish a standard graph for a driver line. Can you trust them?

[littlesaint]

noooo, in parallel = halved (if you've got two equal impedance items wired in parallel). Series doubles.

C

It's technically not halved. It's the inverse of the sum of the inverse of the individual impedances. Though from the math, if they are equal it is indeed halved.

8 and 8 in parallel is 1/( 1/8 + 1/8 ) = 4
8 and 4 in parallel is 1/( 1/8 + 1/4 ) = 2.67

[Dennis H]

Most drivers come with an impedance graph. Can you go by this graph or do you measure your drivers individually. I have a pair of Hippi@#@ something tweeters, and I heard that they were tested and matched individually before leaving the factory. However, most other manufacturers publish a standard graph for a driver line. Can you trust them?

For designing a box, you can probably get close enough with the mfr's specs. For designing a crossover, you should measure the SPL and impedance curves with the drivers mounted in the box as the box will change both curves.

[cjd]

It's technically not halved. It's the inverse of the sum of the inverse of the individual impedances. Though from the math, if they are equal it is indeed halved.

8 and 8 in parallel is 1/( 1/8 + 1/8 ) = 4
8 and 4 in parallel is 1/( 1/8 + 1/4 ) = 2.67

That's why I specified that the impedances must match.

[crazybastard]

I'm thinking about getting the new parts express woofer tester. I'm pretty sure it measures impedance as well. Will it work with mid/tweets as well?